Skip to content

Instantly share code, notes, and snippets.

@m77so

m77so/marsToTSV.ts

Created July 24, 2018 06:52
Show Gist options
  • Save m77so/2c5a193f1222246a96a8c533d188eaa4 to your computer and use it in GitHub Desktop.
Save m77so/2c5a193f1222246a96a8c533d188eaa4 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
marsToTSV.ts
import { OutputJSON } from './dataInterface'
const fs = require('fs')
const iconv = require('iconv-lite')
const path = require('path')
if (process.argv.length - 2 !== 3) {
console.warn('The number of arguments is less or much.\n\nnode marsToTSV.js mars_sd.dat mars_nn.dat outputDir')
process.exit(100)
}
const marsSdDat = process.argv[2]
const marsNnDat = process.argv[3]
const outputDir = process.argv[4]
if (!fs.existsSync(marsSdDat)) {
console.warn(`mars_sd.dat file "${marsSdDat}" does not exist.`)
process.exit(100)
}
if (!fs.existsSync(marsNnDat)) {
console.warn(`mars_nn.dat file "${marsNnDat}" does not exist.`)
process.exit(100)
}
if (!(fs.existsSync(outputDir) && fs.statSync(outputDir).isDirectory())) {
console.warn(`output directory ${outputDir} does not exist.`)
process.exit(100)
}
const output: OutputJSON = {
lineNames: [],
stationNames: [],
lines: [],
stations: [],
cities: [],
appendixFare: {
JRHkansen: { km: [], fare: [] },
JRQkansen: { km: [], fare: [] },
JRSkansen: { km: [], fare: [] }
}
}
const dataSD = fs.readFileSync(marsSdDat)
/**
* 半角カタカナを全角ひらがなに変換
* https://qiita.com/hrdaya/items/291276a5a20971592216
* @param {String} str 変換したい文字列
*/
const hankana2zenkana = function(str: string) {
const kanaMap: { [key: string]: string } = {
ｶﾞ: 'が',
ｷﾞ: 'ぎ',
ｸﾞ: 'ぐ',
ｹﾞ: 'げ',
ｺﾞ: 'ご',
ｻﾞ: 'ざ',
ｼﾞ: 'じ',
ｽﾞ: 'ず',
ｾﾞ: 'ぜ',
ｿﾞ: 'ぞ',
ﾀﾞ: 'だ',
ﾁﾞ: 'ぢ',
ﾂﾞ: 'づ',
ﾃﾞ: 'で',
ﾄﾞ: 'ど',
ﾊﾞ: 'ば',
ﾋﾞ: 'び',
ﾌﾞ: 'ぶ',
ﾍﾞ: 'べ',
ﾎﾞ: 'ぼ',
ﾊﾟ: 'ぱ',
ﾋﾟ: 'ぴ',
ﾌﾟ: 'ぷ',
ﾍﾟ: 'ぺ',
ﾎﾟ: 'ぽ',
ｱ: 'あ',
ｲ: 'い',
ｳ: 'う',
ｴ: 'え',
ｵ: 'お',
ｶ: 'か',
ｷ: 'き',
ｸ: 'く',
ｹ: 'け',
ｺ: 'こ',
ｻ: 'さ',
ｼ: 'し',
ｽ: 'す',
ｾ: 'せ',
ｿ: 'そ',
ﾀ: 'た',
ﾁ: 'ち',
ﾂ: 'つ',
ﾃ: 'て',
ﾄ: 'と',
ﾅ: 'な',
ﾆ: 'に',
ﾇ: 'ぬ',
ﾈ: 'ね',
ﾉ: 'の',
ﾊ: 'は',
ﾋ: 'ひ',
ﾌ: 'ふ',
ﾍ: 'へ',
ﾎ: 'ほ',
ﾏ: 'ま',
ﾐ: 'み',
ﾑ: 'む',
ﾒ: 'め',
ﾓ: 'も',
ﾔ: 'や',
ﾕ: 'ゆ',
ﾖ: 'よ',
ﾗ: 'ら',
ﾘ: 'り',
ﾙ: 'る',
ﾚ: 'れ',
ﾛ: 'ろ',
ﾜ: 'わ',
ｦ: 'を',
ﾝ: 'ん',
ｧ: 'ぁ',
ｨ: 'ぃ',
ｩ: 'ぅ',
ｪ: 'ぇ',
ｫ: 'ぉ',
ｯ: 'っ',
ｬ: 'ゃ',
ｭ: 'ゅ',
ｮ: 'ょ',
'｡': '。',
'､': '、',
ｰ: 'ー',
'｢': '「',
'｣': '」',
'･': '・'
}
const reg = new RegExp('(' + Object.keys(kanaMap).join('|') + ')', 'g')
return str
.replace(reg, match => kanaMap[match])
.replace(/ﾞ/g, '゛')
.replace(/ﾟ/g, '゜')
}
const recordsNumSD = dataSD.length / 28
for (let r = 0; r < recordsNumSD; ++r) {
const offset = 28 * r
let cur = offset
let record = []
record.push(dataSD.readUInt8(cur++))
record.push(dataSD.readInt16LE(cur++))
cur++
record.push(dataSD.readUInt8(cur++))
record.push(iconv.decode(dataSD.slice(cur, cur + 14), 'cp932').replace(/ /g, ''))
cur += 15
record.push(iconv.decode(dataSD.slice(cur, cur + 9), 'cp932').replace(/ /g, ''))
if (!output.lines[record[0]]) {
output.lines[record[0]] = {
id: record[0],
name: record[3],
kana: hankana2zenkana(record[4]),
src: '',
dest: '',
stations: [],
stationIds: [],
kms: [],
akms: [],
dupLineStationIds: [],
chiho: false,
shinkansen: record[3].indexOf('新幹線') > -1,
company: [],
mapZairai: []
}
output.lineNames[record[0]] = record[3]
} else if (record[0] > 0) {
if (output.stationNames.includes(record[3])) {
const id = output.stationNames.indexOf(record[3])
output.stations[id].lineIds.push(record[0])
const lineIds = output.stations[id].lineIds
for (let lineId of lineIds) {
if (!output.lines[lineId].dupLineStationIds.includes(id)) {
output.lines[lineId].dupLineStationIds.push(id)
}
}
} else {
const nextId = output.stations.length
output.stations.push({
id: nextId,
name: record[3],
kana: hankana2zenkana(record[4]),
lineIds: [record[0]],
company: [],
city: -1
})
output.stationNames.push(record[3])
}
const line = output.lines[record[0]]
const stationId = output.stationNames.indexOf(record[3])
line.stations.push(record[3])
line.kms.push(record[1])
line.stationIds.push(stationId)
}
}
const dataNN = fs.readFileSync(marsNnDat)
const recordsNum = dataNN.length / 8
for (let r = 0; r < recordsNum; ++r) {
const offset = 8 * r
let cur = offset
let record = []
record.push(dataNN.readInt16LE(cur))
cur += 2
record.push(dataNN.readInt16LE(cur))
cur += 2
record.push(dataNN.readInt16LE(cur))
cur += 2
record.push(dataNN.readInt16LE(cur))
const id = output.lines[record[0]].kms.indexOf(record[1])
output.lines[record[0]].akms[id] = record[2]
}
for (let line of output.lines) {
const index = [...Array(line.stations.length).keys()].sort((a, b) => line.kms[a] - line.kms[b])
const kms = index.map(i => line.kms[i])
const stations = index.map(i => line.stations[i]) || ['']
const stationIds = index.map(i => line.stationIds[i]) || [-1]
const akms = index.map(i => line.akms[i] || 0)
line.kms = kms
line.stations = stations
line.stationIds = stationIds
line.akms = akms
line.src = stations[0]
line.dest = stations[stations.length - 1]
}
fs.writeFileSync(
path.join(outputDir, 'stations.tsv'),
output.stations.map(s => [s.id, s.name, s.kana].join('\t')).join('\n')
)
fs.writeFileSync(path.join(outputDir, 'lines.tsv'), output.lines.map(s => [s.id, s.name, s.kana].join('\t')).join('\n'))
let distanceArr = []
for (let l of output.lines) {
for (let i = 0; i < l.stationIds.length; ++i) {
distanceArr.push([l.id, l.stationIds[i], l.kms[i], l.akms[i]].join('\t'))
}
}
fs.writeFileSync(path.join(outputDir, 'distances.tsv'), distanceArr.join('\n'))
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment